负载与队列高效结合的主动队列管理算法研究  

Research on Combine of Load-based and Queue-based AQM Algorithm

作　　者：胡闽[1] 李汉菊[1] 

机构地区：[1]华中科技大学计算机科学与技术学院,武汉430074

出　　处：《武汉理工大学学报》2005年第12期89-92,共4页Journal of Wuhan University of Technology

基　　金：国家网络与信息安全保障持续发展计划(2004研1-917-021);国家科技部创新基金(2003(375))

摘　　要：目前已有的大多数主动队列管理算法按照判别拥塞的主要依据可以分为2大分支:基于负载(Load-based)的AQM算法和基于队列(Queue-based)的AQM算法。分析了单独以队列或者单独以负载作为拥塞判别依据的不足,提出了一种基于负载与基于队列相结合的AQM算法LQC(Load Queue Control)算法。仿真结果证明,与RED、FRED和LDC算法相比,LQC算法能更好地稳定队列长度和减少丢包率。The existing AQM algorithms can be devided into two classes according to the indicator which is used to determine its packet drop probabilities： load-based and queue-based AQM algorithms. The disadvantage of the AQM algorithms which use queue length or load alone to indicate congestion degree was analyzed; a kind of AQM algorithm LQC（load queue control） that uses both load and queue length to indicate congestion degree was proposed. Through ns-2 simulation, LQC was compared with RED, FRED and LDC. The simulation results show that LQC can achieve stable queue length and decrease packet loss rate effectively.

关 键 词：AQM(主动队列管理) 拥塞控制 队列长度 丢包率 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]